Behavioral Neuroscience © 2009 American Psychological Association

2009, Vol. 123, No. 3, 624 – 630 0735-7044/09/$12.00 DOI: 10.1037/a0015455

A Behavioral Analysis of the Role of CA3 and CA1 Subcortical Efferents

During Classical Fear Conditioning

Michael R. Hunsaker, Giang T. Tran, and Raymond P. Kesner

University of Utah

It has been proposed that the hippocampus and subcortical structures interact during the processing of
fear and anxiety-related information. It has been demonstrated that the subcortical efferents from CA3
and CA1 can be selectively disrupted without concomitant disruption to the afferents. The present
experiment was designed to evaluate the role of CA3 efferents via the fimbria and the CA1 efferents via
the dorsal fornix for encoding and consolidation/retrieval of classical fear conditioning. The present data
suggest that the subcortical projections from CA3 and CA1 are differentially involved in the processing
of classical fear conditioning, with CA3 subcortical efferents supporting acquisition of both cued and
contextual fear but only supporting retention of contextual fear and CA1 subcortical efferents supporting
the encoding and retrieval of both cued and contextual fear. These data further suggest that all
hippocampal efferents are not homogeneous, and that the hippocampus and subcortex interact to process
conditioned fear.

Keywords: dorsal fornix, fimbria, hippocampus, septum, classical fear conditioning

It has been proposed that septum and hippocampus interact to
modulate anxiety and fear (Gray & McNaughton, 2000; Sheehan,
Chambers, & Russell, 2004). The effects of lesions within the
hippocampal-septal system are mixed—with some reports empha-
sizing an interaction between the hippocampus and septal nuclei,
and others suggesting that the hippocampus and septal nuclei act
independently (Bannerman, Matthews, Deacon, & Rawlins, 2001;
Conejo, Lopez, Cantora, Gonzalez-Pardo, Lopez, & Begega, 2005;
Gray & McNaughton, 1983; Maren & Fanselow, 1997; Phillips &
LeDoux, 1995; Sparks & LeDoux, 1995; Thomas, Yadin, &
Strickland, 1991; Vouimba, Garcia, & Jaffard, 1999). In addition,
in studies of the direct interaction between the hippocampus and
septum, full fimbria/fornix lesions are usually employed, which pre-
clude analysis of the independent contributions of afferent and effer-
ent pathways (cf. Maren & Fanselow, 1997; Phillips & LeDoux,
1995). Despite these difficulties, it has been clearly demonstrated
that the medial and lateral septum process conditioned fear
(Calandreau, Jaffard, & Desmendt, 2007), but they do so in qual-
itatively different ways from the contributions of the dorsal or
ventral hippocampus (Bannerman et al., 2001; Hunsaker &
Kesner, 2008). To date, the role of the hippocampal-septal effer-
ents via the fimbria and dorsal fornix has not been specifically
evaluated for fear-related information processing.
Michael R. Hunsaker, Giang T. Tran, and Raymond P. Kesner, Depart-
ment of Psychology, University of Utah.
Michael R. Hunsaker is now affiliated with Program in Neuroscience,
University of California, Davis.
This research was supported by NSF Grant: IBN-0135273 and NIH
Grant: R01MH065314 awarded to R. P. K. We thank Adam Bevan and
Arik Hone for collecting pilot data.
Correspondence concerning this article should be addressed to Raymond P.
Kesner, Department of Psychology, University of Utah, 380 South 1530 East,
Room 502, Salt Lake City, UT 84112. E-mail: rpkesner@behsci.utah.edu
624
CA3 subcortical efferents project via the fimbria to the lateral
septum, medial septum, and diagonal band of Broca. CA1 subcor-
tical efferents project via the dorsal fornix to the lateral septum,
medial septum, and diagonal band of Broca, as well as to the
mammillary bodies and other structures along the Papez circuit
(Raisman, Cowan, & Powell, 1966; Swanson & Cowan, 1977;
Wyss, Swanson, & Cowan, 1980). More specifically, CA3 projects
to the septofimbrial nucleus, throughout the rostral lateral septum,
the dorsal portion of the medial septum, as well as the dorsal-most
portions of the vertical limb of the diagonal band of Broca. CA1
projects to the septofimbrial nucleus, the ventral portions of the
medial septum, and the horizontal limb of the diagonal band of
Broca. There are only sparse projections to the ventral-most por-
tion of the caudal lateral septum. These differential targets of CA3
and CA1 subcortical efferents suggest that there may be functional
heterogeneity between these two sets of efferents (cf. Hunsaker,
Tran, & Kesner, 2008; Sheehan et al., 2004).
To investigate the roles of these two subcortical efferent path-
ways for the encoding and retrieval of classical fear conditioning,
we have developed a transection paradigm whereby we selectively
disrupt hippocampal efferents to the septum while leaving the
afferents from the septum intact. In this experiment, one group of
animals had CA3 subcortical efferents in the fimbria transected,
but afferents left intact. A second group of animals had CA1
subcortical efferents in the dorsal fornix similarly transected (cf.
Hunsaker, Allan, & Kesner, 2007; Hunsaker, Rogers, & Kesner,
2007; Hunsaker et al., 2008). This experiment was designed to
better characterize the interaction between hippocampal efferent
pathways and their septal targets during tests of learning and
memory. After transecting these pathways, animals were run on a
classical fear-conditioning paradigm to evaluate the effects of the
transections for the encoding and consolidation/retrieval of tone-
cued and contextual fear.
 DORSAL FORNIX AND FIMBRIA IN FEAR CONDITIONING
Materials and Method
Subjects
Sixteen male Long–Evans rats, approximately 4 months of age
and weighing 280 to 350 g at the start of the experiment served as
subjects. The colony room was maintained on a 12-hr light/dark
cycle. All rats had ad libitum access to water and maintained at
90% to 95% of their free feeding weight. All experimental proto-
cols conformed to Institutional Animal Care and Use Committee
(IACUC) and Assessment and Accreditation of Laboratory Animal
Care (AAALAC) regulations. Before participating in the present
experiment, all animals had participated in a study involving a
modified Hebb Williams maze (Hunsaker et al., 2008).
Surgery
Male Long–Evans rats received a transection of CA3 subcortical
efferents in the fimbria (n ⫽ 5), a transection of CA1 subcortical
efferents in the dorsal fornix (n ⫽ 5), or a control surgery (n ⫽ 6;
fimbria control n ⫽ 3; dorsal fornix control n ⫽ 3). The transection
protocols used for these same animals have been detailed else-
where (Hunsaker et al., 2007a,b, 2008). Briefly, animals were
anesthetized with a mixture of ketamine and xylazine (65 mg/kg;
10 mg/kg) and for dorsal fornix transections twisted bipolar stim-
ulating/recording electrodes were lowered into the medial septum
(AP ⫹ 0.7, ML 0.1, DV ⬃4 –5—all at a slight angle away from the
midline (⬃5 degrees) to avoid rupturing any vasculature along
bregma) and CA1 (AP ⫺3.0, ML 1, DV ⬃1.5–2). A ground screw
was secured in the skull above the rostral cortex in lowered to
come in contact with dura. After dura was carefully punctured to
allow easy penetration of electrodes, potentials were alternately
evoked in the medial septum from CA1 and in CA1 from the
medial septum. The placements of the electrodes were adjusted to
maximize this signal in both directions, and then remained un-
touched throughout the remainder of the surgical procedure. A
retractable wire knife was lowered stereotactically until it was
situated adjacent to the dorsal fornix (AP ⫺2.3, ML 0.5, DV ⬃2),
after which the blade was protracted such that it was between CA1
and the dorsal fornix in the alveus. The knife was slowly raised
until the medial septal responses evoked by CA1 stimulation were
attenuated. It was then verified that stimulation of the medial
septum was able to evoke responses in CA1. This was then
performed on the other hemisphere, resulting in a bilateral tran-
section of the dorsal fornix. The choice of the location of the dorsal
fornix transection was such that there was never damage to CA1 or
to the fimbria, also such that the transection was as rostral as
possible without risking disruption of fibers from CA3 to the
fimbria. Control surgery involved everything except the actual
transection (i.e., the knife was lowered into place, protracted, and
then immediately retracted again). For partial fimbria transections
electrodes were lowered into the lateral septum (AP ⫹ 0.6, ML
0.6, DV ⬃3– 4) and CA3 (AP ⫺3.0, ML 3.0, DV ⬃3.0). After dura
was carefully punctured to allow easy penetration of electrodes,
potentials were alternately evoked in the lateral septum from CA3
and in CA3 from the lateral septum. The placements of the
electrodes were adjusted to maximize this signal in both directions,
and then remained untouched throughout the remainder of the
surgical procedure. The retractable wire knife was lowered to be
just ventral to the fimbria (AP 2.3, ML 3.2, DV ⬃4), after which
625
the blade was protracted such that potentials in the lateral septum
from CA3 stimulation were attenuated but the potentials in CA3
from lateral septal stimulation were unchanged. It was then veri-
fied that stimulation of the lateral septum was able to evoke
responses in CA3. This was then performed on the other hemi-
sphere, resulting in a bilateral transection of the fimbria. The
location of the partial fimbria transaction was chosen such that
CA3, the lateral septum, thalamic nuclei, and the dorsal fornix
remained undisturbed.
After surgery, animals were placed in their cages on a heating
pad for 1 to 2 hours to recover and given sweetened, crushed food,
as well as given acetaminophen in their water (Children’s Tylenol;
50 ml/100 ml water) for 3 days after surgery as an analgesic.
Animals were given 2 weeks to recover from surgery before
starting any experimentation.
Behavioral Apparatus and Procedure
Experimental apparatus. Two chambers were used during 3
days of testing. The first chamber was used for encoding and for
the contextual fear retention test. One chamber (28 ⫻ 21 ⫻ 22 cm;
Coulbourn Instruments; Allentown, PA) consisted of two transpar-
ent Plexiglas walls (rear wall and front door) and two aluminum
sidewalls. The chamber floor was made up of 18 steel rods
connected to a shock delivery apparatus used to deliver the shock.
The chamber was surrounded by five small dolls and there were
posters on the walls to provide a clear context. Before and after
conditioning, the chamber was cleaned with an unscented cleaning
solution to remove any olfactory cues an animal may have left
during conditioning.
The second observation chamber was used during tests of
auditory-cued fear retention. This chamber (32 cm2) was con-
structed entirely of transparent Plexiglas. No cues surrounded this
chamber, only a green curtain and the table upon which the
chamber was set. The floor of the chamber was covered with a
single paper towel before placing the rat inside. Before and after
conditioning, the chamber was cleaned with the same unscented
cleaning solution used in the other chamber to remove olfactory
cues.
Behavioral methods
Encoding—Day 1. Rats were placed in the conditioning cham-
ber for 2 minutes before the first auditory stimulus, after which
they received 10 auditory-shock pairings separated by 74 seconds.
An auditory stimulus (10-s duration, 2 kHz, 85-dB) was presented
through a speaker to initiate each trial. An electric foot shock (2-s
duration, 0.75 mA) was presented coterminal with the auditory
stimulus. A 74-s intertrial interval (ITI) separated each successive
tone ⫹ shock pairing. After 10 tone ⫹ shock pairings and subse-
quent ITIs, rats remained in the chamber for an additional 2
minutes. A freezing response was measured by an observer who
scored freezing behavior every 8 seconds during the preexposure
period and ITI (the 2 seconds after the shock were discarded and
freezing during the subsequent 72 seconds was recorded) and
every 4 seconds during the tone stimulus; resulting in two auditory
observations and nine ITI observations for each trial. The first nine
trials of acquisition were blocked into 3-trial blocks for analysis of
acquisition. Tone and contextual (ITI) freezing were also blocked
626 HUNSAKER, TRAN, AND KESNER
into a single 10-trial block for further characterization of any
specific tone or contextual fear acquisition deficits. It is understood
that the contextual measure is not completely dissociable from
tone-related freezing, but previous research has shown that sepa-
rating tone and contextual freezing as we have here is a fair
approximation of separate processes (Hunsaker & Kesner, 2008).
Contextually fear retention test—Day 2 or 3. Each rat was
tested for retrieval of contextually cued fear either 24 or 48 hours
after acquisition (half on Day 2 and half on Day 3, counterbalanced
with auditory-cued fear retention tests). The rat was placed in the
encoding chamber for 8 minutes in the absence of the auditory cue
and shock for eight minutes (blocked for analysis into eight, 1-min
trials). Freezing behavior was measured every 8 seconds. Because
of extinction in the control group, only the first 4 minutes of testing
were blocked and used for statistical analysis.
Auditory-cued fear retention test—Day 2 or 3. Each rat was
tested for retrieval of auditory-cued fear either 24 or 48 hours after
acquisition (half on Day 2 and half on Day 3, counterbalanced with
contextual fear retention tests). The rat was placed in the encoding
chamber for 8 minutes in the continuous presence of the auditory
cue for 8 minutes (blocked for analysis into eight, 1-min trials).
Freezing behavior was measured in 8-s intervals. Because of
extinction in control animals, only the first 4 minutes of testing
were blocked and used for statistical analysis.
Histology
At the conclusion of testing, each rat was deeply anesthetized
with an injection of sodium pentobarbital (70 mg/kg i.p.), perfused
intracardially with PBS and 10% (wt/vol) formalin, and the brain
was removed from the skull and stored at 4 °C in a solution of 30%
sucrose in 10% formalin (wt/vol) for 72 hours before further
processing. The brain was frozen and 40 ␮m sections collected
through the medial septum and hippocampus for Nissl and acetyl-
cholinesterase staining.
Dependent Measures and Statistical Analysis
Freezing scores during encoding and retrieval tests were trans-
formed to percentages and were blocked for statistical analysis. A
two-way repeated measures ANOVA with groups as the between
factor and blocks of three trials as the within factor was employed
for testing group differences during the encoding of delay fear
conditioning (e.g., the first nine trials of acquisition were blocked
into three, 3-trial blocks). The last trial was excluded arbitrarily to
keep all blocks the same size and to keep the number of blocks at
three, as has been presented previously (Hunsaker & Kesner,
2008). One-way ANOVA with groups as the between factor were
used to analyze encoding (single 10-trial acquisition block during
acquisition) and retention of contextual and auditory-cued fear
(single 4-min block during the tests) between groups. All main
effects were considered significant at p ⬍ .05. Fisher’s least
significant difference (LSD) post hoc paired comparisons tests
were performed upon all significant effects. Overall acquisition of
fear conditioning for each group is presented in three blocks of
trials, plotted as means ⫾SE. Contextual and auditory-cued fear
encoding were collapsed into a single 10-trial block for analysis.
Retention tests are presented as mean percent freezing ⫾SE for
single 4-min blocks.
Results
Histology
The results of the partial fimbria and dorsal fornix transections
are depicted in Figure 1. More thorough anatomical, neurophysi-
ological, and neurochemical analyses of the lesions of these same
animals have been reported elsewhere (Hunsaker et al., 2007,
2008). In short, for animals with dorsal fornix transections, the
responses in the medial septum from CA1 stimulation were re-
duced by the transection, whereas the CA1 responses from medial
septal stimulation were unchanged. For animals with partial fim-
bria transections, the responses in the lateral septum from CA3
stimulation were reduced by the transection, whereas the CA3
responses from the lateral septum were unchanged. Anatomical
tract tracing experiments using biotynilated dextran amine previ-
ously reported by Hunsaker et al. (2008) revealed that the dorsal
fornix transection disrupted ⬃50% of CA1 subcortical efferents,
without significantly disrupting projections from the medial sep-
tum into the hippocampus. It should be noted that since only 50%
of these efferents were transected, the dorsal fornix transection was
not complete. Partial fimbria transections disrupted ⬃63% of CA3
subcortical efferents, without significantly disrupting projections
from the septum into the hippocampus. Again, it must be empha-
sized that this manipulation was incomplete. In the case of either
transection, there was no disruption in acetylcholinesterase stain-
ing in the hippocampus after the transection, additionally suggest-
ing that there was no large-scale disruption to the projections from
the medial septum into the hippocampus (Figure 1A–C). Detailed
anatomical, neurophysiological, and histochemical data that dem-
onstrate the specificity and efficacy of these manipulations have
been reported previously (cf. Hunsaker et al., 2007a,b, 2008). The
reader is referred to those reports since all the animals in the
present study were those used by Hunsaker et al. (2008).
Behavior
Figure 2A shows the overall acquisition of classical fear condi-
tioning. A two-way repeated measures ANOVA with groups (fim-
bria transection, dorsal fornix transection, control) as the between
factor and blocks of 3-trials as the within repeated factor were
Figure 1. Histology. (A) Acetylcholinesterase (AChE) stained section
from a partial fimbria transection. (B) AChE stained section from a control
animal. (C) AChE stained section from a dorsal fornix transection. There
were no differences in AChE reactivity between control animals and
animals receiving either of the transections. In both (A) and (C) the
transections are signaled by arrows. (D) Diagram of a partial fimbria
transection after (A). (E) Diagram of a dorsal fornix transection. In both
(D) and (E) the largest (gray) and smallest (black) lesions are depicted. The
reader is referred to Hunsaker et al. (2008) for more detailed neurophysi-
ological, anatomical, and histochemical analyses of these lesions.
 DORSAL FORNIX AND FIMBRIA IN FEAR CONDITIONING 627

Figure 2. Encoding and Consolidation/Retrieval of delay fear conditioning. (A) Both the partial fimbria
transection group and the dorsal fornix transection groups showed overall acquisition deficits relative to control
animals. (B) The partial fimbria transection and dorsal fornix transected animals showed deficits for the encoding
of tone and contextual fear during acquisition. ⴱ p ⬍ .05. ⴱⴱ p ⬍ .01. (C) The partial fimbria transection group
showed no differences from controls for retention of tone-cued fear, whereas the dorsal fornix transected animals
showed large deficits. Both the partial fimbria and dorsal fornix transected animals showed deficits relative to
controls for the retention of contextual fear. ⴱ p ⬍ .05. ⴱⴱ p ⬍ .01.

performed on the data as plotted. There was a significant group
effect, F(1, 15) ⫽ 12.00, p ⬍ .002, an effect for blocks, F(2, 30) ⫽
6.55, p ⬍ .006, and a group ⫻ block interaction, F(2, 30) ⫽ 7.87,
p ⬍ .001. A Fisher’s LSD post hoc paired comparisons test
revealed that both the fimbria transection and dorsal fornix tran-
section groups showed an overall acquisition deficit relative to
control animals (all ps ⬍ 0.01). Fimbria transected animals and
dorsal fornix transected animals did not differ ( p ⬎ .1). Despite the
deficit in acquisition, the transection groups did show some acqui-
sition and were not at floor levels. It also should be noted that no
animals appeared either hyperactive or showed freezing behaviors
during the 2-min baseline period before acquisition, though quan-
titative data were not collected.
To evaluate whether it was a specific deficit in tone or contex-
tual fear encoding, the acquisition data were separated into tone
encoding and contextual encoding (Figure 2B), and the separated
data were collapsed into a single 10-trial block for analysis. One-
way ANOVA with group as the between factor were used to
analyze tone and contextual fear encoding. For tone acquisition,
there was an effect for group, F(2, 15) ⫽ 45.98, p ⬍ .001. A
Fisher’s LSD post hoc paired comparison revealed that both the
fimbria transection and dorsal fornix transection group froze less
than the control group ( ps ⬍ 0.01). The fimbria transection and
dorsal fornix transection groups did not differ ( p ⬎ .1). For context
acquisition, there was an effect for groups, F(2, 15) ⫽ 96.17, p ⬍
.001. A Fisher’s LSD post hoc paired comparisons test revealed
that the dorsal fornix transection groups froze less than the control
group ( p ⬍ .01). The dorsal fornix transection group froze less
than the fimbria transection group ( p ⬍ .05). The fimbria transec-
tion group also froze less than the control group ( p ⬍ .05). This
suggests the fimbria transection group showed a lesser impairment
relative to the dorsal fornix transection group. Additionally,
though the dorsal fornix group showed impairments in acquisition
relative to the other groups, they were not at baseline freezing
levels. Their learning was impaired relative to the other groups, but
not absent.
One-way ANOVA with group as the between factor were used
to evaluate the tone and contextual fear retention data. The tone
and contextual fear data were blocked into single 4-min blocks for
analysis (Figure 2C). For the retention of tone-cued fear, there was
628 HUNSAKER, TRAN, AND KESNER
an effect for groups, F(2, 15) ⫽ 64.97, p ⬍ .001. A Fisher’s LSD
post hoc paired comparisons test revealed that the dorsal fornix
transection group froze less than both the fimbria transection group
and controls ( ps ⬍ 0.01). Fimbria transected animals and control
animals froze similarly ( p ⬎ .1). For the retention of contextual
fear, there was an effect for groups, F(2, 15) ⫽ 96.25, p ⬍ .001.
A Fisher’s LSD post hoc paired comparisons test revealed that
fimbria transected animals froze less than control animals ( p ⬍
.01), but froze more than dorsal fornix transected animals ( p ⬍
.01). Dorsal fornix transected animals froze less then control
animals, as well ( p ⬍ .01). This pattern of results suggests that the
fimbria transection group showed a lesser impairment relative to
the dorsal fornix transection group.
Discussion
The present data suggest that although the CA3 and CA1 effer-
ents into the lateral and medial septum (and CA1 efferents to the
subcortex) both participate in the acquisition of conditioned fear,
they do so in a slightly different manner. This is reflected in the
fact that animals without CA3 efferents to the lateral septum have
some consolidation/retrieval of contextual fear and efficient con-
solidation/retrieval auditory-cued fear, whereas animals without
CA1 efferents appear to have little to no memory whatsoever of
the conditioned fear. Although the animals without CA1 subcor-
tical efferents (the dorsal fornix transection group) as well as the
animals without CA3 subcortical efferents (the partial fimbria
transection group) showed deficits for acquisition, they were not at
floor and did appear to acquire some cued fear, just less efficiently
than the control group. This is reflected in the fact that the animals
froze more during conditioning than they did during the 2-min
baseline condition and the first trial of conditioning.
Although counterintuitive, it has been demonstrated that encod-
ing (or acquisition) and consolidation/retrieval processes are par-
tially dissociable (cf. Aggleton & Saunders, 1997; Vann &
Aggelton, 2004). In fact, recent studies have shown similar phe-
nomena in primates and rodents with lesions to the fornix (Aggle-
ton & Saunders, 1997; Buckley, Eilson, & Gaffan, 2008; Eacott &
Gaffan, 2005; Mitchell, Browning, Wilson, Baxter, & Gaffan,
2008) and in studies with lesions along the Papez circuit (targets of
the projections in the fornix; cf. Vann & Aggleton, 2004). In fact,
the same fimbria transection animals that participated in the
present experiment previously caused acquisition or encoding def-
icits on a modified Hebb-Willliams maze, without concomitant
deficits for retrieval (i.e., they had difficulty learning the task, but
remembered what they had learned as well as the control group;
Hunsaker et al., 2008). That is not to say that encoding and
retrieval are completely dissociable, but it means that there can be
intact retrieval after impaired acquisition, and vice versa (cf.
Hunsaker et al., 2008).
It has been proposed that the hippocampus and subcortical
structures form a neural circuit that modulates anxiety in the
animal (Gray & McNaughton, 2000). It has also been determined
that there are unique patterns of behavioral and cognitive deficits
after selective lesions at different locations within the septal-
hippocampal circuit (Gray & McNaughton, 1983). The model
proposed by Gray and McNaughton (2000) posits dynamic, real-
time interactions between the hippocampus and septum during
some learning and memory paradigms. There are reports of medial
septal lesions resulting in deficits for classical fear conditioning
(Bannerman et al., 2001; Maren & Fanselow, 1997; Sparks &
LeDoux, 1995; cf. Calandreau et al., 2007 for an inactivation
experiment), but the present data do not match these previous
findings. It has also been demonstrated that fornix lesions result in
deficits for fear conditioning (Phillips & LeDoux, 1995). The
present deficit pattern does not mimic those seen after dorsal and
ventral CA1 and CA3 excitotoxic lesions (Hunsaker & Kesner,
2008), nor do they mimic the effects of full dorsal or ventral
hippocampal lesions (Bannerman et al., 2001). These data suggest
that the current transections disrupted an interaction between the
hippocampus and the subcortex, and did not result in specific
septal or hippocampal deficits per se. Further lesion and inactiva-
tion studies are necessary to better characterize the hippocampal-
subcortical interactions disrupted by these selective transections.
In previous reports (Hunsaker et al., 2007a, 2008), two animals
received biotinylated dextran amine (BDA; an anterograde tracer)
infusions 7 days postunilateral partial fimbria transection and two
animals had BDA injections 7 days postdorsal fornix transection.
The results suggested that the CA3 and CA1 subcortical efferents
project to separate populations within the septal nuclei (the study
did not investigate other subcortical structures aside from the
septal nuclei). After infusions of BDA into CA3, there was stain-
ing present in the septofimbrial nucleus, throughout the rostral
lateral septum, the dorsal portion of the medial septum, as well as
the dorsal-most portions of the vertical limb of the diagonal band
of Broca. After infusions of BDA into CA1, there was staining in
the septofimbrial nucleus, the ventral portions of the medial sep-
tum, and the horizontal limb of the diagonal band of Broca. There
was only sparse staining in the ventral-most portion of the caudal
lateral septum. The differential targets of CA3 and CA1 subcorti-
cal efferents suggest that there may be functional heterogeneity
between these two sets of efferents (cf. Hunsaker et al., 2008;
Sheehan et al., 2004). Additionally, during this anatomical tract
tracing experiment, it was demonstrated that the transections of the
fimbria disrupted roughly 63% of CA3 efferent fibers the dorsal
fornix transections disrupted slightly more than 50% of the CA1
efferent fibers.
Unlike previous reports by Hunsaker et al. (2007a,b, 2008) that
posited the subcortical outputs of the hippocampus were primarily
modulatory on the hippocampus via a feed forward inhibition
system, the present data suggest that information processing within
the medial and lateral septum may be directly affected by efferents
from the hippocampus during the conditioning of auditory-cued
and contextual fear (cf. Gray & McNaughton, 2000; Sheehan et al.,
2004). Any differences between the effects of transecting CA1 or
CA3 subcortical efferents may be attributed to the differences in
the specific targets of the projections. It has previously been
demonstrated that there are dissociations between the medial and
lateral septum for auditory-cued and contextual fear processing
(Calandreau et al., 2007). The present data suggest that the CA3
efferents (primarily) to the lateral septum are involved for the
acquisition of contextual fear, involved in consolidation/retrieval
of contextual fear, but only involved for the acquisition of
auditory-cued fear (i.e., the transection left auditory-cued fear
consolidation/retrieval intact). Also of note is that the retrieval
deficits were never as severe as those displayed by the dorsal
fornix transection group. The present data also suggest that the
CA1 efferents (primarily) to the medial septum (and structures
 DORSAL FORNIX AND FIMBRIA IN FEAR CONDITIONING
along the Papez circuit) are involved during acquisition and (po-
tentially) for the consolidation/retrieval of auditory-cued and con-
textual fear, though the present data are unclear on the nature of the
retrieval deficits, though these data are congruent with previous
findings (Phillips & LeDoux, 1995). The present data do not lend
themselves to a specific analysis of encoding as fully separable
from consolidation or retrieval processing as well as previous
studies (i.e., there needs to be some sort of measurable consolida-
tion/retrieval to determine whether some learning had taken
place—such as that seen in the animals with partial fimbria tran-
sections; cf. Hunsaker & Kesner, 2008; Hunsaker et al., 2008).
An alternative explanation for the deficits seen after dorsal
fornix transactions may be that subcortical efferents from the
subiculum were also disrupted by the dorsal fornix transactions—
not just those from CA1. These subcortical efferents from the
subiculum also have been shown to project not only to the medial
septum, but also to structures along the Papez circuit (Gray &
McNaughton, 2000; Swanson & Cowan, 1977; Vann & Aggleton,
2004), which may also have been involved.
The present data add to the understanding of the neural circuitry
underlying affective and fear related learning, as well as provide
further insight into the subcortical projections of CA1. What the
present data suggest is that the accepted view of the hippocampus
supporting contextual and the amygdala mediating both contextual
and elemental (e.g., auditory) fear conditioning is at best incom-
plete (cf. Quinn, Wied, Ma, Tinsley, & Fanselow, 2008). These
findings also suggest that using fear conditioning as a test of
amygdala function may be confounded by the presence of
hippocampal-septal (and hippocampal-subcortical) interactions
(cf. Conejo et al., 2005). Additionally, these data suggest that
hippocampal and septal interactions are involved during fear-
related learning and not just during anxiety-related processing
(Gray & McNaughton, 2000). Of particular interest is the finding
that disrupting subcortical efferents from either CA1 or CA3 was
sufficient to disrupt acquisition of classically conditioned fear.
Retrieval of conditioned contextual fear was dependent upon sub-
cortical projections from CA3 (the data do not allow any mean-
ingful interpretations for CA1 subcortical efferents other than that
they are critical for acquisition). The present study reveals a role
for subcortical structures in both hippocampus dependent and
(presumably) hippocampus independent processing of conditioned
fear.
In the case of the present experiment, it appears that the hip-
pocampus and subcortical structures interact to efficiently process
the auditory and contextual cues during the acquisition and re-
trieval of classical fear conditioning. This suggests that the medial
septum, in addition to sending cholinergic and GABAergic pro-
jection fibers into the hippocampus may potentially play an im-
portant role in the formation of a representation of the tone and
context that are associated with an aversive shock. The lateral
septum, it appears, also may have a role in the formation and recall
of these representations. Additional subcortical effects of the tran-
sections may potentially contribute to deficits seen after the dorsal
fornix transections. The medial septum has strong, reciprocal
connectivity with the entorhinal cortex and the amygdala, and any
alteration in the outputs from the septum to these areas may also
potentially underlie the present effects (Gray & McNaughton,
2000; Sheehan et al., 2004; Swanson & Cowan, 1977; Vann &
Aggleton, 2004). Overall, these data suggest that the role of the
629
hippocampal subcortical efferents and their role in learning and
memory processing should be reevaluated.
References
Aggleton, J. P., & Saunders, R. C. (1997). The relationships between
temporal lobe and diencephalic structures implicated in anterograde
amnesia. Memory, 5, 49 –71.
Bannerman, D. M., Matthews, P., Deacon, R. M., & Rawlins, J. N. (2001).
Medial septal lesions mimic effects of both selective dorsal and ventral
hippocampal lesions. Behavioral Neuroscience, 118, 1033–1041.
Buckley, M. J., Eilson, C. R., & Gaffan, D. (2008). Fornix transection
impairs visuospatial memory acquisition more then retrieval. Behavioral
Neuroscience, 122, 44 –53.
Calandreau, L., Jaffard, R., & Desmedt, A. (2007). Dissociated roles for the
lateral and medial septum in elemental and contextual fear conditioning.
Learning and Memory, 14, 422– 429.
Conejo, N. M., Lopez, M., Cantora, R., Gonzalez-Pardo, H., Lopez, L.,
Begega, A., et al. (2005). Effects of pavlovian fear conditioning on
septohippocampal metabolism in rats. Neuroscience Letters, 373, 94 –
98.
Eacott, M. J., & Gaffan, D. (2005). The roles of perirhinal cortex, postrhi-
nal cortex, and the fornix for objects, contexts, and events in the rat.
Quarterly Journal of Experimental Psychology, B, 58, 202–217.
Gray, J. A., & McNaughton, N. (1983). Comparison between the behav-
ioural effects of septal and hippocampal lesions: A review. Neuroscience
and Biobehavioral Reviews, 7, 119 –188.
Gray, J. A., & McNaughton, N. (2000). The neuropsychology of anxiety:
An enquiry into the functions of the septo-hippocampal system. Oxford,
UK: Oxford University Press.
Hunsaker, M. R., Allan, K. D., & Kesner, R. P. (2007a). The role of CA3
efferents via the fimbria for the acquisition of a delay non-match to place
task. Hippocampus, 17, 494 –502.
Hunsaker, M. R., & Kesner, R. P. (2008). Dissociations across the dorsal-
ventral axis of CA3and CA1 for encoding and retrieval of contextual and
auditory-cued fear. Neurobiology of Learning and Memory, 89, 61– 69.
Hunsaker, M. R., Rogers, J. L., & Kesner, R. P. (2007b). Behavioral
characterization of a transection of dorsal CA3 subcortical efferents:
Comparison with scopolamine and physostigmine infusions into dorsal
CA3. Neurobiology of Learning and Memory, 88, 127–136.
Hunsaker, M. R., Tran, G. T., & Kesner, R. P. (2008). Dissociating
subcortical hippocampal efferents for encoding and consolidation/
retrieval of spatial information: CA1 from CA3. Hippocampus, 18,
699 –709.
Maren, S., & Fanselow, M. S. (1997). Electrolytic lesions of the fimbria/
fornix, dorsal hippocampus, or entorhinal cortex produce anterograde
deficits in contextual fear conditioning in rats. Neurobiology of Learning
and Memory, 67, 142–149.
Mitchell, A. S., Browning, P. G., Wilson, C. R., Baxter, M. G., & Gaffan,
D. (2008). Dissociable roles for cortical and subcortical structures in
memory retrieval and acquisition. Journal of Neuroscience, 28, 8387–
8396.
Phillips, R. G., & LeDoux, J. E. (1995). Lesions of the fornix but no the
entorhinal or perirhinal cortex interfere with contextual fear condition-
ing. Journal of Neuroscience, 15, 5308 –5315.
Quinn, J. J., Wied, H. M., Ma, Q. D., Tinsley, M. R., & Fanselow, M. S.
(2008). Dorsal hippocampus involvement in delay fear conditioning
depends upon the strength of the tone-footshock association. Hippocam-
pus, 18, 640 – 654.
Raisman, G., Cowan, M. W., & Powell, T. P. S. (1966). An experimental
analysis of the efferent projection of the hippocampus. Brain, 89, 83–
108.
Sheehan, T. P., Chambers, R. A., & Russell, D. S. (2006). Regulation of
affect by the lateral septum: Implications for neuropsychiatry. Brain
Research Reviews, 46, 71–117.
630 HUNSAKER, TRAN, AND KESNER

Sparks, P. D., & LeDoux, J. E. (1995). Septal lesions potentiate freezing
behavior to contextual but not phasic conditioned stimuli in rats. Behav-
ioral Neuroscience, 109, 184 –188.
Swanson, L. W., & Cowan, M. W. (1977). An autoradiographic study of
the organization of the efferent connections of the hippocampal forma-
tion in the rat. Journal of Comparative Neurology, 172, 49 – 84.
Thomas, E., Yadin, E., & Strickland, C. E. (1991). Septal unit activity
during classical conditioning: A regional comparison. Brain Research,
547, 302–308.
Vann, S. D., & Aggleton, J. P. (2004). The mammillary bodies: Two
memory systems in one? Nature Reviews: Neuroscience, 5, 35– 44.
Vouimba, R. M., Garcia, R., & Jaffard, R. (1999). Pretraining tetanic
fimbrial stimulation impairs the expression but not the acquisition of
contextual fear conditioning in mice. Neuroscience, 93, 869 – 876.
Wyss, X., Swanson, L. W., & Cowan, M. W. (1980). The organization of
the fimbria, dorsal fornix, and ventral hippocampal commissure in the
rat. Anatomy and Embryology, 138, 303–316.
Received February 22, 2008
Revision received January 8, 2009
Accepted January 18, 2009 䡲

Members of Underrepresented Groups:

Reviewers for Journal Manuscripts Wanted
If you are interested in reviewing manuscripts for APA journals, the APA Publications and
Communications Board would like to invite your participation. Manuscript reviewers are vital to the
publications process. As a reviewer, you would gain valuable experience in publishing. The P&C
Board is particularly interested in encouraging members of underrepresented groups to participate
more in this process.

If you are interested in reviewing manuscripts, please write to APA Journals at Reviewers@apa.org.
Please note the following important points:

• To be selected as a reviewer, you must have published articles in peer-reviewed journals. The
experience of publishing provides a reviewer with the basis for preparing a thorough, objective
review.

• To be selected, it is critical to be a regular reader of the five to six empirical journals that are most
central to the area or journal for which you would like to review. Current knowledge of recently
published research provides a reviewer with the knowledge base to evaluate a new submission
within the context of existing research.

• To select the appropriate reviewers for each manuscript, the editor needs detailed information.
Please include with your letter your vita. In the letter, please identify which APA journal(s) you
are interested in, and describe your area of expertise. Be as specific as possible. For example,
“social psychology” is not sufficient—you would need to specify “social cognition” or “attitude
change” as well.

• Reviewing a manuscript takes time (1– 4 hours per manuscript reviewed). If you are selected to
review a manuscript, be prepared to invest the necessary time to evaluate the manuscript
thoroughly.